-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
4.14.z
-
None
-
Moderate
-
None
-
3
-
ETCD Sprint 256, ETCD Sprint 257, ETCD Sprint 258
-
3
-
False
-
Description of problem:
Upgrading a fresh ("healthy") OCP < 4.14.18 cluster to 4.14.24 results in almost doubling of the etcd_network_peer_round_trip_time_seconds_bucket metric from +/- 15ms to +/- 25ms.
Version-Release number of selected component (if applicable):
upgrade from 4.14.18 and 4.14.24
How reproducible:
Steps to Reproduce:
1. install new cluster <= 4.14.18
2. etcd_network_peer_round_trip_time_seconds_bucket value is +/- 15ms
3. upgrade to 4.14.24
4. etcd_network_peer_round_trip_time_seconds_bucket value is doubled to +/- 25ms
Actual results:
Expected results:
Additional info:
